Regulatory T cells control PF4/heparin antibody production in mice

Abstract

Heparin-induced thrombocytopenia (HIT) is a relatively common drug-induced immune disorder that can have life-threatening consequences for affected patients. Immune complexes consisting of heparin, platelet factor 4 (PF4) and PF4/heparin-reactive antibodies are central to the pathogenesis of HIT. Regulatory T (Treg) cells are a subpopulation of CD4 T cells that play a key role in regulating immune responses, but their role in controlling PF4/heparin-specific antibody production is unknown. In the studies described here, we found that FoxP3-deficient mice lacking functional Treg cells spontaneously produced PF4/heparin-specific antibodies. Following transplantation with bone marrow cells from FoxP3-deficient but not wild-type mice, Rag1-deficient recipients also produced PF4/heparin-specific antibodies spontaneously. Adoptively transferred Treg cells prevented spontaneous production of PF4/heparin-specific antibodies in FoxP3-deficient mice and inhibited PF4/heparin complex-induced production of PF4/heparin-specific IgGs in wild-type mice. Treg cells suppress immune responses mainly through releasing anti-inflammatory cytokines, such as interleukin-10 (IL-10). IL-10-deficient mice spontaneously produced PF4/heparin-specific antibodies. Moreover, BM chimeric mice with CD4 T cell-specific deletion of IL-10 increased PF4/heparin-specific IgG production upon PF4/heparin complex challenge. Short-term IL-10 administration suppresses PF4/heparin-specific IgG production in wild-type mice. Taken together, these findings demonstrate that Treg cells play an important role in suppressing PF4/heparin-specific antibody production.

Introduction

Heparin-induced thrombocytopenia (HIT) is the most common drug-induced, immune-mediated thrombocytopenia, usually occurring 3–6 days following heparin administration (1, 2). Antibodies specific for platelet factor 4 (PF4) in a complex with heparin or a similar polyanion are a hallmark of HIT (1-3) and react with PF4/heparin to form ultra-large immune complexes that are central to HIT pathogenesis (4, 5). When formed in vivo, these complexes may directly activate FcγRIIa on the platelet surface to induce platelet activation (6, 7). Alternatively, activation via FcγRIIa may result from antibody recognition of small quantities of PF4 bound to the platelet surface (8, 9). In either case, platelet activation leads to thrombocytopenia and a high risk of arterial and/or venous thrombosis or thromboembolism, ischemic limb necrosis, pulmonary embolism, myocardial infarction and stroke (6, 7).

Breakdown of B cell tolerance can trigger PF4/heparin-specific antibody production in a mouse model (10). Regulatory T (Treg) cells are a subset of circulating CD4 T cells that are critical for the maintenance of peripheral immune tolerance, especially in prevention of autoimmunity and immunopathology during immune responses (11). A previous report showed that Treg function can be broadly inhibited by PF4 and/or PF4/heparin complexes and suggested that this might contribute to HIT pathogenesis (12). The transcription factor forkhead box 3 (FoxP3) is a master regulator of Treg cells and is essential for Treg cell lineage specification during development and necessary for Treg suppressive function (13). Treg cells control immune reactions through production of immnuno-regulatory cytokines, such as TGF-β, IL-10 and IL-35 (14, 15). Deficiency of functional Treg cells caused by mutations of FoxP3 leads to spontaneous systemic multi-organ auto-inflammatory phenotypes in Scurfy mice (16, 17) and immune dysregulation, polyendocrinopathy, enteropathy and X-linked (IPEX) syndrome in humans (18). Here, we show that Treg cells play an important role in PF4/heparin-specific antibody production in a mouse model.

Materials and Methods

Mice

Foxp3^ΔEGFP (FoxP3- deficient) and Foxp3^EGFP mice were the gift of Dr. Talal Chatila (Harvard University). CD4-deficient and IL-10-deficient mice on a C57BL/6 genetic background, wild-type C57BL/6, and Rag1-deficient mice were purchased from Jackson Laboratory. Experimental and control mice were 8- to 10-week-old. Mice were maintained in the Biological Resource Center at the Medical College of Wisconsin. All animal protocols were approved by the MCW Institutional Animal Care and Use Committee.

PF4/heparin administration

Administration of mouse PF4/heparin complexes daily to mice with a C57BL/6 genetic background for 5 days has been described previously (19-21). Briefly, mouse PF4 and heparin were mixed in 1 x Hanks balanced salt solution (Invitrogen) at final concentrations of 200 μg/ml of PF4 and 4 U/ml of heparin, a previously optimized molar ratio (2.6:1) for PF4/heparin complex formation and immunogenicity, followed by 1-hour incubation at room temperature. Mice anesthetized with isoflurane were injected retro-orbitally, with 100 μl of the mouse PF4/heparin solution once daily for 5 days. Following injection, sera were collected at 7-day intervals. Mouse PF4/heparin-specific antibodies were measured by ELISA. Specificity of the assay was confirmed by experiments showing that antibodies produced display greater binding to mouse PF4/heparin than to mouse PF4 alone and that antibody binding is inhibited by an excess amount of heparin.

Bone marrow (BM) transplantation

BM cells were harvested from wild-type or FoxP3-deficient mice and then transplanted into sub-lethally irradiated (600 rads) 8- or 10-week-old Rag1-deficient mice by intravenous injection (5 × 10⁶ cells/recipient). Three weeks later, sera were collected from the recipients for measuring the levels of mouse PF4/heparin-specific antibodies.

BM chimeric mice were generated to study the effect of CD4-specific deficiency of IL-10 on the production of PF4/heparin-specific antibodies. BM cells from CD4-deficient mice were mixed with BM cells from wild-type or IL-10-deficient mice in PBS supplemented with 2% FBS at a 7:3 ratio. The mixed BM cells were transplanted into lethally irradiated (1000 rads) 8- or 10-week-old CD4-deficient mice by intravenous injection (5 × 10⁶ cells/recipient). Eight weeks later, the BM chimeric recipients, in which CD4 T cells were wild-type or IL-10-deficient and others cell were largely normal, were injected with mouse PF4/heparin complexes. Sera were collected at the indicated time points for measurement of the levels of mouse PF4/heparin-specific antibodies.

Adoptive transfer

For the rescue of newborn FoxP3-deficient mice, 1 × 10⁶ Treg cells (CD4⁺EGFP⁺) combined with either 4 × 10⁶ CD4⁺EGFP^– conventional T cells, as a source of induced Treg (iTreg) cells or 1×10⁶ in vitro derived iTreg cells were injected into the peritoneal cavity of Foxp3^ΔEGFP (FoxP3-deficient) pups within the first 30 hours after birth as previously described (22). Eight to 10 weeks after the neonatal rescue, sera were collected from the recipients and the levels of mouse PF4/heparin-specific antibodies were measured.

For the Treg treatment of WT mice. CD4⁺EGFP⁺ Treg cells were sorted from the lymph nodes and spleens of 8-week-old CD45.1⁺ Foxp3^EGFP mice. Cells were suspended in PBS supplemented with 2% FBS and then intravenously injected into 8- or 10-week-old wild-type mice (0.5 × 10⁶/recipient). One hour after adoptive transfer, the wild-type recipients were administered with mouse PF4/heparin complexes daily for 5 days and sera collected at the indicated time points.

Flow cytometry

Spleens or inguinal lymph nodes were harvested from wild-type mice that were adoptively transferred with conventional T cells (CD45.1⁺CD4⁺EGFP⁻) or Treg cells (CD4⁺EGFP⁺). Single-cell suspensions of splenocytes or lymph node cells were treated with Gey solution to remove red blood cells and were resuspended in phosphate-buffered saline (PBS) supplemented with 2% fetal bovine serum (FBS). The cells were then stained with a combination of fluorescence-conjugated antibodies. Allophycocyanin (APC)-conjugated anti-CD4 was purchased from eBioscience. APC-Cy7-conjugated anti-CD45.1 was purchased from BioLegend. Samples were applied to a flow cytometer (LSRII, Becton Dickinson). Data were collected and analyzed using FACSDiva software (Becton Dickinson).

P-selectin expression assay

Platelet activation was measured by the P-selectin expression assay (23). Washed platelets (1.5 × 10⁶) from wild-type or transgenic mice expressing human FcγRIIa on platelets were incubated at a concentration of 30 × 10⁶/ml with mouse PF4 (20 μg/ml) in the presence of low dose (0.02 U/ml) or high dose (20 U/ml) heparin for 20 minutes at room temperature. Platelets incubated with 1% BSA served as a negative control. Sera from wild-type or FoxP3-deficient mice were then added and the mixture was incubated for an additional 1 hour. Platelets stimulated with 0.5 μM Phorbol 12-myristate 13-acetate (PMA) served as a positive control. Subsequently, platelets were stained with FITC-conjugated anti-CD41/61 (αIIbβ3; EMFRET Analytics, #M025–1) and PE-conjugated anti-CD62P (P-selectin; eBioscience, #12– 0626-80) antibodies for 20 minutes at room temperature. P-selectin expression on CD41/61-positive platelets was measured by flow cytometry.

Inhibition of PF4/heparin-specific antibody production by IL-10 administration

Wild-type mice were intraperitoneally injected with recombinant mouse IL-10 (PeproTech, 2.5 μg/mouse) or PBS one hour before PF4/heparin immunization. After once-daily injection and immunization for 5 days, sera were collected at a 7-day interval and PF4/heparin-specific antibodies were measured by ELISA.

Statistical analysis

Statistical analysis was performed with two-tailed unpaired Student’s t test for the samples with normal distribution or Mann-Whitney test for the samples with random distribution. To determine the differences among three groups of samples, a one-way ANOVA followed by Tukey’s post-hoc test or a repeated measures one-way ANOVA followed by two-tailed paired Student’s t test was performed. The normality of all the data was analyzed by D’Agostino & Pearson omnibus normality test using Graphpad Prism software.

Results

FoxP3-deficient mice spontaneously produce PF4/heparin-specific antibodies

B-cell tolerance plays an important role in controlling the production of PF4/heparin-specific antibodies in mice (10). Treg cells are critical for immune tolerance, and lack of functional Treg cells due to FoxP3 deficiency results in severe autoimmunity in both humans and mice (16-18). To investigate the role of Treg cells in PF4/heparin-specific antibody production, we examined whether FoxP3-deficient mice were prone to produce these antibodies. FoxP3-deficient mice become moribund at about 4 weeks of age (24), thus we measured the levels of PF4/heparin-specific antibodies in sera from 3-week-old mutant mice and their littermate controls. Unmanipulated FoxP3-deficient mice produced markedly higher levels of PF4/heparin-specific IgMs (Fig. 1A) and IgGs (Fig. 1B) than wild-type controls. These IgMs and IgGs reacted more strongly with PF4/heparin complexes than with PF4 alone, and the reaction was inhibited by excess heparin (Fig. 1C, 1D) as is typical of antibodies from patients with HIT. Nonetheless, these IgMs and, particularly, the IgGs reacted only weakly with PF4 alone, consistent with studies showing that FoxP3-deficient mice develop antibodies to various self-antigens (16-18). Thus, unmanipulated FoxP3-deficienct mice can spontaneously produce PF4/heparin-specific antibodies.

Figure 1.

FoxP3 deficiency results in spontaneous production of PF4/heparin-specific antibodies. FoxP3-deficient mice spontaneously produce PF4/heparin-specific antibodies. Sera from unmanipulated 3-week-old wild-type (FoxP3^+/+) or FoxP3-deficient (FoxP3^−/−) mice were collected and examined for the levels of mouse PF4/heparin-specific IgMs (A) or IgGs (B) by ELISA. Specificity of the IgMs (C) or IgGs (D) toward mouse PF4, mouse PF4/heparin, and mouse PF4 in the presence of excess heparin was examined. Data shown are obtained from 9 wild-type and 11 FoxP3-deficient mice. Statistical analysis was performed with two-tailed unpaired Student’s t test (A,B) and a repeated measures one-way ANOVA followed by a two-tailed paired Student’ s t test (C,D).

A generally accepted criterion for pathogenicity of an HIT antibody is that it be able to activate platelets expressing the IgG FcγRII receptor CD32 in the presence of PF4 or low-dose heparin in a reaction that is inhibited by high-dose heparin (4, 25). To examine whether PF4/heparin-specific antibodies from unmanipulated FoxP3-deficient mice exhibited this property, we examined their ability to induce activation of PF4-treated platelets. In the presence of mouse PF4 and low-dose heparin, sera from FoxP3-deficient but not wild-type mice activated mouse platelets from transgenic mice expressing human FcγRIIa but not wild-type platelets that lacked FcγRIIa, leading to expression of P-selectin (Fig. 2). In contrast, sera from FoxP3-deficient mice failed to activate FcγRIIa-expressing platelets treated with BSA rather than PF4 (Fig. 2). In addition, the ability of sera from FoxP3-deficient mice to activate FcγRIIa-expressing platelets was strongly inhibited by high dose heparin (Fig. 2). Of note, PMA activated platelets with or without human FcγRIIa to express P-selectin (Fig. 2). These data demonstrate that sera from FoxP3-deficient mice are able to activate platelets in an FcγRIIa- and PF4-dependent manner, indicating PF4/heparin-specific antibodies produced by FoxP3-deficienct mice are functionally active.

Figure 2.

FcγRIIa- and mPF4/heparin-dependent platelet activation by sera from FoxP3-deficient mice. Platelets from wild-type (FcγRIIa⁻) or human FcγRIIa transgenic mice (FcγRIIa⁺) were incubated with mPF4 in the presence of low dose (mPF4 + Lo Hep) or high dose (mPF4 + Hi Hep) heparin. Then sera from wild-type (FoxP3^+/+) or FoxP3-deficient (FoxP3^−/−) mice were added. Platelets incubated with BSA served as negative controls and platelets stimulated with PMA served as positive controls. Platelet activation was measured by expression of P-selectin. The expression of P-selectin on CD41/61-positive platelets was determined by flow cytometry. The numbers indicate mean fluorescence intensity (MFI). Data shown are representative (A) or summarized (B) data from 6 wild-type and 7 FoxP3-deficient mice. Statistical analysis was performed with two-tailed unpaired Student’s t test.

We next examined whether spontaneous production of PF4/heparin-specific antibodies by FoxP3-deficient mice is BM cell-intrinsic and transplantable. BM cells from FoxP3-deficient or wild-type controls were transplanted into sub-lethally irradiated Rag1-deficient mice that have no functional T and B cells (26). Twenty one days after transplantation, recipients that received FoxP3-deficient BM cells produced higher levels of PF4/heparin-specific IgMs than recipients that received wild-type control BM cells (Fig. 3A). In addition, recipients that received FoxP3-deficient but not wild-type BM cells spontaneously produced PF4/heparin-specific IgGs (Fig. 3B). These IgGs and IgMs reacted more strongly with PF4/heparin complexes than with PF4 alone, and binding was inhibited by excess heparin (data not shown). These findings show that FoxP3 deficiency-induced spontaneous production of PF4/heparin-specific antibodies is BM cell-intrinsic and transplantable.

Figure 3.

Rag1-deficient mice transplanted with FoxP3-deficient BM spontaneously produce mouse PF4/heparin-specific antibodies. Partially irradiated Rag1-deficient mice were transplanted with BM cells from wild-type (Rag1^−/− + FoxP3^+/+) or FoxP3-deficient (Rag1^−/− + FoxP3^−/−) mice. Sera were collected from the recipients before (Day 0) and 21 days after (Day 21) transplantation, and examined for the levels of mouse PF4/heparin-specific IgMs (A) or IgGs (B) by ELISA. Each dot represents a mouse, and the horizontal lines indicate the mean values. Data shown are obtained from 10 recipients of each type. Statistical analysis was performed with two-tailed unpaired Student’s t test (A) and Mann-Whitney test (B).

Adoptive transfer of Treg cells suppresses PF4/heparin-specific antibody production

Autoimmunity in FoxP3-deficient mice is due to the lack of functional Treg cells (24). We therefore examined whether spontaneous production of PF4/heparin-specific antibodies is regulated by Treg cells. FACS-sorted Treg cells from Foxp3^EGFP mice that express enhanced green fluorescent protein (GFP) in FoxP3⁺ Treg cells (CD4⁺EGFP⁺) were adoptively transferred into the peritoneal cavity of newborn FoxP3-deficient mice. FoxP3-deficient mice lacking transferred Treg cells spontaneously produced PF4/heparin-specific IgM and IgG antibodies compared to FoxP3-sufficient mice (Fig. 4A and B). In contrast, FoxP3-deficient mice with adoptively transferred Treg cells failed to produce PF4/heparin-specific IgMs and IgGs (Fig. 4A, 4B). These data show that adoptively transferred Treg cells can prevent spontaneous production of PF4/heparin-specific antibodies in FoxP3-deficient mice.

Figure 4.

Adoptively transferred Treg cells prevent and IL-10 deficiency increases production of PF4/heparin-specific antibodies, respectively. (A,B) Adoptively transferred Treg cells prevent spontaneous production of mouse PF4/heparin-specific antibodies in FoxP3-deficient mice. Treg cells (EGFP⁺) were sorted from Foxp3^EGFP mice and injected into the peritoneal cavity of newborn FoxP3-deficient mice. Sera were collected 3-week after injection, and examined for the levels of mouse PF4/heparin-specific IgMs (A) or IgGs (B) by ELISA. (C) Adoptively transferred Treg cells prevent PF4/heparin complex administration-induced production of mouse PF4/heparin-specific antibodies in wild-type mice. Conventional CD4 T cells (CD4⁺EGFP⁻) or Treg cells (EGFP⁺) were sorted from FoxP3^EGFP mice and intravenously injected into wild-type mice. Following administration with mouse PF4/heparin complexes, sera were collected at the indicated time points after administration. The levels of mouse PF4/heparin-specific IgGs were measured by ELISA. Data shown are obtained from 5 wild-type mice without adoptive transfer (FoxP3^+/+ + none), 5 FoxP3-deficeint mice without adoptive transfer (FoxP3^−/− + none) and 6 FoxP3-deficient with adoptive transferred Treg cells (FoxP3^−/− + Treg) mice (A,B), from 8 recipients of Tconv group and 10 recipients of Treg group (C). Statistical analysis was performed with a one-way ANOVA followed by Tukey’s post-hoc test (A,B) or two-tailed unpaired Student’s t test (C).

We also examined the role of Treg cells in controlling PF4/heparin complex-induced production of PF4/heparin-specific antibodies. Treg cells (CD4⁺EGFP⁺) or conventional CD4 T cells (CD4⁺EGFP⁻) were sorted from Foxp3^EGFP mice and then adoptively transferred into wild-type mice, followed by PF4/heparin complex administration. It is known that PF4/heparin complex challenge mainly induces PF4/heparin-specific IgGs but not IgM (19). As expected, wild-type mice that received conventional CD4 T cells produced PF4/heparin-specific IgG at day 7 after challenge, and the levels of these IgGs gradually declined thereafter (Fig. 4C). In contrast, PF4/heparin-specific IgGs production was significantly reduced in wild-type recipients that received Treg cells (Fig. 4C). We have shown that adoptively transferred Treg cells are functional (27). Consistently, 21 days after the adoptive transfer, the transferred conventional CD4 T cells or Treg cells were detected in the recipients (Fig. 5). Taken together, these data demonstrate that transferred Treg cells suppress production of PF4/heparin-specific antibodies in wild-type mice.

Figure 5.

Detection of both Tconv and Treg cells in the recipients after adoptive transfer. Wild-type mice were adoptively transferred with CD45.1⁺CD4⁺EGFP⁻ conventional T cells (WT + Tconv) or CD4⁺EGFP⁺ Treg cells (WT + Treg). The presence of adoptively transferred Tconv and Treg cells in the spleen and inguinal lymph nodes of the recipients were detected by flow cytometry in the CD4-gated populations 21 days after transfer.

Treg cells inhibit PF4/heparin-specific antibody production partially through IL-10

Treg cells suppress immune responses mainly through releasing anti-inflammatory cytokines, including interleukin-10 (IL-10) (14). We examined the role of IL-10 in production of PF4/heparin-specific antibodies. Unmanipulated IL-10-deficient mice spontaneously produced significantly higher levels of PF4/heparin-specific IgMs and IgGs than wild-type controls (Fig. 6A, 6B). These IgG and IgM antibodies strongly bound to PF4/heparin complexes but not PF4 alone (data not shown). Thus, IL-10 deficiency results in spontaneous production of PF4/heparin-specific antibodies.

Figure 6.

IL-10 plays an important role in PF4/heparin-specific antibody production. (A,B) Spontaneous production of mouse PF4/heparin-specific antibody in IL10-deficient mice. Sera were collected from unmanipulated 8- to 10-week-old wild-type (IL10^+/+) or IL10-deficient (IL10^−/−) mice and examined for the levels of mouse PF4/heparin-specific IgMs (A) or IgGs (B) by ELISA. (C) Increased production of mouse PF4/heparin-specific antibodies in CD4 T cell-specific IL-10-deficient chimeric mice upon PF4/heparin administration. BM cells from CD4-deficient mice were mixed with BM cells from wild-type (CD4^−/− + IL-10^+/+) or IL10-deficient (CD4^−/− + IL-10^−/−) mice at a 7:3 ratio, and then transplanted into lethally irradiated CD4-deficient mice. BM chimeric mice were administered with mouse PF4/heparin complexes 8 weeks after transplantation. Sera were collected at the indicated time points after administration, and the levels of mouse PF4/heparin-specific IgGs were measured by ELISA. (D,E) IL-10 suppresses PF4/heparin-specific antibody production in wild-type mice. Wild-type mice were intraperitoneally injected with mouse IL-10 or PBS one hour before PF4/heparin immunization. After once-daily injection and immunization for 5 days, sera were collected at a 7-day interval and PF4/heparin-specific IgM (D) and IgG (E) were measured by ELISA. Each dot represents a mouse, and the horizontal lines indicate the mean values. Data shown are obtained from 11 wild-type and 10 IL10-deficient mice (A,B), from 5 BM chimeric mice with wild-type CD4 T cells and 8 BM chimeric mice with IL10-deficient CD4 T cells (C), and 10 wild-type mice in each group (D,E). Statistical analysis was performed with two-tailed unpaired Student’s t test.

CD4⁺ T cells, including Treg cells, are an important source of IL-10 production (28). In addition, a variety of other types of cells, such as epithelia, activated macrophages, dendritic cells and B cells, produce IL-10 (28). Thus, we examined the effect of IL-10 deficiency in CD4 T cells on PF4/heparin-specific antibody production. We generated CD4 T cell-specific IL-10-deleted BM chimeric mice by transplanting mixed BM cells from IL-10-deficient and CD4-deficient mice into lethally irradiated CD4-deficient mice. The resulting BM chimeric mice possessed CD4 T cells that were derived from IL-10-deficient mice. CD4-specific deletion of IL-10 alone resulted in slight, but not significant, spontaneous production of PF4/heparin-specific antibodies in unchallenged BM chimeric mice (Fig. 6C). However, following immunization with PF4/heparin complexes, BM chimeric mice with IL-10-deficient CD4 T cells produced markedly higher levels of PF4/heparin-specific IgGs than control BM chimeric mice that received a mixture of wild-type and CD4-deficeint BM cells and therefore possessed wild-type CD4 T cells (Fig. 6C). Of note, PF4/heparin complex challenge mainly induces PF4/heparin-specific IgG (19). In addition, both groups of BM chimeric mice had a normal population of Treg cells (data not shown). Treg cells control immune responses by secreting several anti-inflammatory cytokines, such as IL-10, TGF-β and IL-35 (15). These data indicate that Treg cells regulate PF4/heparin-specific antibody production at least partially through production of the anti-inflammatory cytokine IL-10.

Furthermore, we examined whether IL-10 administration could suppress PF4/heparin-specific antibody production. Wild-type mice were injected with mouse IL-10, followed by PF4/heparin complex challenge. Importantly, IL-10 suppressed the production of PF4/heparin-specific IgGs but not IgMs in wild-type mice (Fig. 6D, 6E). However, IL-10 administration failed to inhibit the production of PF4/heparin-specific IgMs or IgGs in IL-10-deficient mice (data not shown). The fact that IL-10 can suppress PF4/heparin-specific IgG production in wild-type mice indicates a therapeutic potential of IL-10 for preventing and/or treating HIT.

Discussion

HIT is caused by antibodies reactive against a cryptic epitope expressed on PF4 when it is in a complex with heparin or another similarly charged macromolecule (29). Both PF4 and heparin are self-antigens, suggesting that loss of self-tolerance could contribute to the production of PF4/heparin-specific antibodies and classifying HIT as an autoimmune-like disorder. In fact, mice have been shown to possess pre-existing, inactive PF4/heparin-specific B cells that can be activated to produce antibodies by in vitro or in vivo pro-inflammatory influences or in vivo deletion of the B-cell tolerance regulator protein kinase Cδ (10). Importantly, healthy humans also possess pre-existing, inactive PF4/heparin-specific B cells that can produce antibodies following in vitro stimulation with pro-inflammatory molecules (10). Thus, breakdown of B-cell immune tolerance plays an important role in anti-PF4/heparin antibody production.

Normally, B-cell tolerance to self-antigens is established or maintained by clonal deletion and receptor editing, in which autoreactive B cells are eliminated, and by anergy that reversibly inactivates autoreactive B cells (30). Dysregulation of B-cell anergy impairs B-cell tolerance, leading to anti-PF4/heparin antibody production in mice (10). Moreover, Treg cells can suppress the immune responses of autoreactive B cells and thus play an important role in controlling peripheral immune tolerance (11). Impaired function or reduced number of Treg cells are related to autoantibody production in human patients (31, 32). In contrast, administration of Treg cells into autoimmune animals reduces autoantibody production (33). Here we demonstrate that Treg deficiency results in development of PF4/heparin-specific antibodies in mice. Of note, Treg-deficient mice spontaneously produce multiple different autoreactive antibodies, including anti-platelet antibodies that induce thrombocytopenia (34). It is not known whether PF4/heparin-specific antibodies exist in the human patients with IPEX syndrome, an extremely rare autoimmune disease and related to defective Treg cells. Nonetheless, PF4/heparin-specific antibodies have been indeed found in the autoimmune patients with systemic lupus erythematosus or antiphospholipid antibodies (35, 36). These findings suggest that extra caution might be needed when administrating heparin in patients with a history of autoimmune disease.

PF4/heparin-specific antibodies from mice are dominantly IgG2b and IgG3 (19). The important question is whether PF4/heparin-specific antibodies produced by FoxP3-deficienct mice are functional. Administration of the hPF4/heparin-specific mouse monoclonal antibodies (KKO) along with heparin induces thrombocytopenia and thrombosis in mice expressing human PF4 and human FcγRIIA transgenes (37). However, the requirement of a large quantity of hPF4/heparin-specific antibodies to induce thrombocytopenia/thrombosis in the mouse model prevents us from examining whether the sera from FoxP3-deficient mice are functional in vivo. Importantly, the sera from FoxP3-deficient mice are able to activate platelets in vitro, indicating PF4/heparin-specific antibodies produced by FoxP3-deficienct mice are functionally active.

Treg cell deficiency leads to autoantibody production (31, 32). Treg-deficient mice can spontaneously produce anti-platelet antibodies that induce thrombocytopenia (34). Thus, it is possible that platelet-autoreactive, but not PF4/heparin-specific, antibodies from FoxP3-deficient mice might be responsible for platelet activation in vitro. It would be ideal to examine whether purified PF4/heparin-specific antibodies from FoxP3-deficient mice can activate platelets in vitro. The maximum amount of sera from a 3-week-old Foxp3-deficient mouse is limited, preventing us from obtaining sufficient purified PF4/heparin-specific antibodies for the in vitro platelet activation assay. Nonetheless, mouse platelets with but not those without a human FcγRIIa transgene can be activated by the sera from FoxP3-deficient mice in the presence of mPF4. In addition, the platelet activation by the sera from FoxP3-deficient mice is inhibited by high dose of heparin. These findings demonstrate that the activation of mouse platelets induced by sera from FoxP3-deficient mice is FcγRIIa-, mPF4- and heparin-dependent. Therefore, PF4/heparin-specific antibodies in the sera from FoxP3-deficient mice are responsible for platelet activation in vitro.

Treg cells usually target T helper (Th) cells to indirectly suppress the B cell response, but can also suppress antibody production and antibody class switching by directly interacting with B cells (38, 39). The lack of functional Treg cells in IPEX patients disrupts peripheral but not central B cell tolerance, resulting an increase of auto-reactive B cells in the mature naive B cell compartment (40). Our findings show that Treg cells regulate PF4/heparin-induced antibody production. Multiple types of cells, such as CD4 T cells, dendritic cells, monocytes and macrophages, are likely involved in the regulation of PF4/heparin-specific antibody production (41-43). It is unclear whether Treg cells directly or indirectly suppress PF4/heparin-specific B cells. Treg cells can exert their functions through the release of inhibitory cytokines, such as IL-10, transforming growth factor β (TGF-β) and IL-35 (44). Our findings suggest that Treg cells regulate PF4/heparin-specific antibody production at least partially through production of IL-10. IL-10-deficient mice display a systemic immune dysregulation (14, 15). Short-term administration of IL-10 is not expected to reverse the systemic immune dysregulation and thus to inhibit PF4/heparin-induced production of PF4/heparin-specific IgGs in IL-10-deficient mice. Importantly, short-term IL-10 injection is able to suppress PF4/heparin-specific IgG production in wild-type mice, confirming a critical role of IL-10 in PF4/heparin-specific antibody production and suggesting a therapeutic potential of IL-10 for preventing and/or even treating HIT. Of note, it has been shown that PBMCs from HIT patients generate lower levels of IL-10 following PF4/heparin stimulation, suggesting a potential role of IL-10 in human HIT pathogenesis (45). Further studies are warranted to investigate the mechanism by which Treg cells regulate PF4/heparin-specific antibody production.

Key Points:

Treg cells regulate PF4/heparin-specific antibody production through IL-10.

Treg cells are important for the pathogenesis of heparin-induced thrombocytopenia.